基于数值模拟的湿润锋前进法测量精度分析

非饱和土的渗透系数函数测量难度大、耗时长，湿润锋前法(WFM)可在短时间内测得渗透系数函数(HF)，但依赖肉眼识别的湿润锋，存在一定的局限性，且该方法的测量精度尚不明确，有待验证。文中针对湿润锋前进法进行研究，探讨初始含水率、湿润锋阀值、降雨入渗速率、传感器位置等因素对湿润锋前进法的测量精度的影响。采用Seep/W软件模拟均质土柱的入渗过程，分析湿润锋前进法数据，计算土体的渗透系数，将其和输入的渗透系数（可认为是真实解）进行比较，对湿润锋前进法的计算精度进行评估，并讨论误差的来源。分析结果表明，湿润锋前进法能够获得比较精确的计算结果；使用湿润锋特征含水率计算湿润锋前进速率，突破了原始湿润锋前进法存在“肉眼观察”的局限性，大大拓展了该方法的适用性；传感器间距对湿润锋前进法渗透系数函数计算精度没有直接影响；初始含水率越低，降雨入渗速率越大，渗透系数函数范围跨度越大。基于文中的分析结果，对湿润锋前进法的试验设计时，建议采用30～50 cm的土柱进行试验，传感器的数量建议为3～4，可以采用任意初始含水率进行试验接近干燥更好，为避免表面积水，建议降雨强度小于饱和渗透系数。

Accuracy analysis of wetting front advancing method based on numerical simulation

The permeability coefficient function of unsaturated soil is difficult to be measured and a long time is required. Wetting front advancing method(WFM) can be applied to measure the hydraulic conductivity function(HF) of unsaturated soil in a short time. However, this method relies on an artifical recognition of wet front, which has certain limitations. In addition, the measurement accuracy of WFM is not clear and needs to be verified. In this paper, impacts of parameters include of initial water content, critical water content of wetting front, rainfall intensity, location of sensor etc on the accuracy of WFM is analyzed. The Seep/W software is used to simulate the infiltration process occurs in the homogeneous soil column. The data is analyzed by WFM to calculate the permeability coefficient of the soil. Simulation results are compared with the input permeability coefficient(i.e. the real solution). The calculation accuracy of the wetting front method is evaluated and the source of error is discussed. The results show that WFM can obtain relative high accurate calculation results. Using the characteristic moisture content at the wetting front to calculate wetting front advancing rate, it overcomes the limitations of original wetting front advancing method with artificial observation and greatly expands the applicability of WFM. The sensor spacing has no direct effect on the calculation accuracy of the permeability coefficient function of wetting frontal advance method. The lower the initial water content, the greater the rainfall infiltration rate, and the greater the span of the permeability coefficient function. Based on the analysis in this paper, the following suggestions are proposed for test design of WFM. The soil column with a length of 30-40 cm and 3-4 sensors are suggested. Any initial water content is acceptable and the dry sample is the best choice. To avoid the surface ponding problem, the rainfall intensity should be smaller than the saturated permeability coefficient of soil.